When the audition room fell silent, neither Arthur nor Tom looked first at the young performer standing beneath the lights.
They looked at the cameras.
Not one camera.
Six.
One broadcast camera, three vertically mounted smartphones streaming to different platforms, one AI-assisted tracking camera, and a laptop continuously generating real-time captions in multiple languages. Before a single line of dialogue had been spoken, software had already estimated the performer’s age range, vocal tempo, emotional intensity, audience retention probability, and even which facial expressions would likely become short video clips.
Arthur smiled.
“So this is rehearsal now.”
Half a century earlier, Arthur had been mocked for insisting that actors memorize exact stage coordinates.
“Stand 3.2 meters from the left wing.”
“Turn exactly thirty degrees.”
“Pause for one-point-five seconds.”
Critics accused him of murdering theater with engineering.
Tom had called it “factory acting.”
Yet Arthur had never wanted mechanical performances.
He simply understood something before television directors themselves fully understood it.
A camera does not see like an audience.
A few centimeters determine whether a face remains inside the frame.
A slight shift changes the focal plane.
Under studio lighting, reproducibility is not the enemy of emotion—it is the infrastructure that allows emotion to be captured consistently.
Decades later, computer vision would formalize many of these principles. Modern virtual production systems, motion-capture pipelines, and automated camera tracking all rely on repeatable spatial data. What Arthur had done with masking tape on a stage anticipated ideas that later became commonplace in digital filmmaking.
Tom had been right, too.
The underground theater movement of the 1970s had inherited the philosophy of the “happening,” where each performance was meant to exist only once. Its value came precisely from resisting repetition. The audience participated in creating an event that could never be perfectly recreated.
Neither man had been entirely wrong.
History had simply changed the question.
⸻
The young applicant finished her monologue.
No one applauded.
Instead, a screen displayed analytics.
Viewer attention: 87%.
Predicted completion rate: 74%.
Emotional engagement peaked at 42 seconds.
Likelihood of viral short-form extraction: High.
Tom frowned.
“When did applause become a graph?”
Arthur answered quietly.
“When audiences stopped sitting in one room.”
⸻
Streaming platforms had altered economics.
Recommendation algorithms rewarded content that retained attention within seconds. Media companies increasingly optimized productions for measurable engagement, while creators adapted performances for clips, thumbnails, and algorithmic distribution. Even live broadcasts were now planned with moments designed to become shareable fragments.
Television personalities had gradually become something else.
Not actors.
Not celebrities.
Not influencers.
All three simultaneously.
Their identities had become intellectual property continuously expanded across television, podcasts, livestreams, merchandise, and social media. Marketing researchers called this a “transmedia personal brand,” where audiences followed a persistent individual across many formats rather than a single fictional character.
Tom remembered what he had once told Arthur.
“People don’t care about someone disappearing into a role anymore.”
Arthur nodded.
“They care about recognizing someone immediately.”
Recognition had become more valuable than transformation.
⸻
The afternoon’s final applicant entered without announcing a character.
He simply introduced himself.
He spoke about repairing motorcycles.
He joked about failing university entrance exams.
He demonstrated a magic trick learned from his grandfather.
Then he read Shakespeare.
Nothing connected.
Everything connected.
The panel watched in silence.
Tom whispered,
“He’s not showing range.”
Arthur replied,
“No.”
“He’s showing identity.”
After the auditions ended, the building emptied except for the two old men.
Tokyo’s skyline shimmered through the glass, giant digital billboards continuously refreshing according to audience demographics detected in real time.
Arthur laughed.
“You know…”
“My obsession with reproducibility lost.”
Tom laughed too.
“My obsession with uniqueness lost.”
Arthur tilted his head.
“Maybe they both won.”
Tom looked outside.
Artificial intelligence could now clone voices.
Generate faces.
Create realistic performances from text.
Digital humans could reproduce expressions with almost perfect consistency.
Ironically, the qualities that machines had become exceptionally good at were exactly the ones over which theater had once argued.
Consistency.
Repeatability.
Precision.
The remaining scarcity was not technical excellence.
It was a believable human presence.
After a long silence, Arthur repeated the invitation he had made decades before.
“Why don’t we work together again?”
Tom smiled.
“This time…”
“…let’s produce people who can survive being copied.”
Arthur raised an eyebrow.
Tom continued.
“Not by performing better than machines.”
“But by remaining unmistakably human, even after every gesture can be reproduced.”
Outside, another livestream began.
Inside, for the first time in fifty years, neither of them argued about theater.
They discussed authentication, synthetic media, digital provenance, and the curious possibility that the oldest purpose of performance—to convince another human being that someone is truly present—had become the newest challenge of the AI era.
All names of people and organizations appearing in this story are pseudonyms
Comments